Scientists in China use AI to pinpoint new therapeutic targets in PAH
Computer-based analyses uncover 2 genes that may drive rare PH type
Written by |
Using artificial intelligence (AI) and analyzing genetic data focused on two primary biochemical processes, a team of scientists in China has identified two genes that appear to play key roles in driving pulmonary arterial hypertension (PAH).
Computer-based analyses suggest that these two genes, called ATP1B1 and HP, may be useful therapeutic targets in PAH — though the scientists cautioned that further work is needed to verify and expand upon these results.
“Machine learning algorithms … were employed to identify the … genes,” the researchers wrote, noting that “this study was designed to identify key hub genes … and investigate their mechanistic roles.”
The goal, ultimately, is to develop new treatments for this rare type of pulmonary hypertension, according to the researchers.
Their study, “Identification of Key Genes Related to SUMOylation and Potassium Channels and Their Mechanism of Influence on Pulmonary Arterial Hypertension,” was published in the journal Biochemical Genetics.
PAH is marked by abnormally high pressure in the vessels that carry blood through the lungs. The causes of PAH are not fully understood, but a few key biochemical processes have been implicated in the disease’s development. For example, previous studies have indicated that PAH is marked by abnormalities with potassium channels, which are proteins that allow the transport of potassium ions across the cell membrane.
Cells use a range of different molecular modifications to control the activity of proteins such as potassium channels. One type of modification that’s especially important for modulating potassium channel activity is known as SUMOylation, short for small ubiquitin-like modifier-mediated modification.
Scientists use AI to zero in on so-called hub genes
One of the hardest parts about studying the underlying biology of PAH and other complex diseases is that, if the activity of one gene changes, there are often dozens or hundreds of other genes whose activity will change as a direct consequence.
Thus, when researchers see that many genes are dysregulated, it can be difficult to zero in on the genes that are originally responsible for the widespread changes. These genes are often referred to as hub genes, because when researchers draw diagrams of how all the genes affect each other, they appear at the center like the hub of a wheel.
Now, taking advantage of publicly-available genetic databases, this research team employed machine learning to identify hub genes specifically related to potassium channels and SUMOylation in PAH. Machine learning is a field of artificial intelligence that basically works by feeding large sets of data to a computer, alongside complex mathematical algorithms that the computer uses to identify patterns in the data.
“The application of machine learning algorithms enables the detection of complex, non-linear relationships between genes that traditional statistical methods might overlook,” the researchers wrote. “By employing an integrated bioinformatic approach to dissect these potential molecular mechanisms, this study was designed to investigate the hub genes at the intersection of SUMOylation and potassium channels in PAH, thereby providing a theoretical basis for the development of novel therapeutic strategies.”
The computational analyses identified five candidate hub genes. To validate the computational data, the researchers then measured the activity of these genes in blood samples from five people with PAH and five people without the disease. The results showed that two of the genes, ATP1B1 and HP, were significantly more active in cells from PAH patients.
The researchers therefore posited that these two genes are likely hub genes involved in driving PAH. Both genes have previously been implicated in maintaining blood vessel health, though they have not been specifically linked to PAH before, the team noted.
Findings shed new light on biology underlying PAH
In further analyses, the researchers noted that increased activity of these two genes was often accompanied by markers of increased inflammation.
“The concomitant upregulation of ATP1B1 and HP in this inflammatory milieu suggests that these hub genes may either contribute to or be amplified by the NF-κB signaling axis,” the scientists wrote. That’s a well-known biological process driving inflammation, according to the team.
The scientists also noted that, based on the computational data, these genes seem to be particularly active in epithelial cells, which are the cells that line the inside of blood vessels.
The team also conducted computer-based analyses to look for existing drugs that may modulate the activity of these two hub genes. Their results pointed to a medication called digoxin, which is already commonly used for PAH, the team noted.
“In clinical practice, digoxin has been extensively utilized in the treatment of PAH patients and is widely considered an effective therapeutic agent,” the researchers wrote.
Overall, the findings of this study shed further light on the underlying biology of PAH and point to potential treatment targets, the team noted. The scientists stressed, however, that their analyses were largely limited to computer-mediated tests of data from small numbers of patients, so further work is needed to validate the findings.
